|

Shopify Product Schema: 7 Fields AI Search Reads

ByTyler Ziton

Shopify product schema is the JSON-LD block on each product page that tells Google and ChatGPT what a product is, costs, and whether it is in stock, and seven fields do almost all of that work, and they are the ones every AI surface reads first. Your theme outputs most of them. Finish the block and cut the duplicates.

Key Points

  • Seven fields carry the load: name, description, price and currency, availability, brand, ratings and reviews, and a GTIN or MPN.
  • Dawn and most modern themes output Product schema already, so the usual problem is two or three competing blocks from apps, not a missing one.
  • Google says no special markup is needed for AI Overviews or AI Mode, but the same fields feed rich results, Shopping, and the ChatGPT product feed.
  • Check the top 20 products once a quarter with the Rich Results Test and the Merchant listings report, then leave it alone.

What Shopify Product Schema Is

Shopify product schema is a block of JSON-LD in the code of a product page that spells out the name, price, stock, brand, and reviews in a format a machine reads without guessing. The visible page is written for a shopper and a crawler struggles to tell the price from the SKU. Google calls it a merchant listing.

Think of that block as the label on the back of the box. The front of the box is the product page a person sees: photos, a headline, a buy button. The label on the back lists the same facts in the order a machine expects, and every store’s label uses the same fields, so Google can compare 10,000 stores in one pass.

That label is what earns you the extras in search. Star ratings under the listing, the price and “In stock” line, and the product carousels in AI answers all pull from it. Whether that markup also earns AI citations is a separate question, and I cover it in does schema help AI citations. The short version is that it removes doubt, and machines rank what they do not have to guess about.

Does Shopify Already Add Product Schema?

Yes, Shopify already adds product schema on every product page in Dawn and the themes built on it, since the theme writes a Product JSON-LD block with the name, image, price, currency, and availability pulled from your product data. A fresh store on a modern theme passes the Rich Results Test before anyone touches code.

The problem starts with apps. A review app injects its own Product block so it can add stars. An SEO app injects a third block to be safe. Now the page carries two or three Product blocks with different values, and Google picks one for you, sometimes the one without your reviews. Duplicate Product blocks are the most common structured data problem I find on a Shopify audit, ahead of missing fields by a wide margin.

So the fix is one block that carries everything. Let the theme own the Product schema. Switch the review app to its integration setting, which feeds ratings into the theme’s block instead of writing a second one. Then turn off the SEO app’s schema if it has one. I go through which apps earn their keep and which ones just add code in my rundown of Shopify SEO apps.

The 7 Product Schema Fields AI Search Reads

The 7 product schema fields AI search reads are name, description, offers, availability, brand, aggregateRating with review, and gtin or mpn. Those are the fields Google’s merchant listing spec and OpenAI’s product feed spec both name. Only three of them are required. The rest are where the clicks come from.

The required three come from Google’s merchant listing documentation, which asks only for name, image, and an offer with a price above zero and a three-letter currency code. Everything else on this list is recommended, and a recommended field is what turns a bare listing into one with stars and a price.

1. name

The name field is the product’s title, written exactly the way it reads on the page. Google matches this against the visible headline, and a mismatch is a policy violation, not a typo. Shopify fills it from the product title, so the only thing to check is that you did not let an app rewrite it with a keyword stuffed in.

So keep the name human. “Ventani Hybrid Heated Jacket” tells a model what the product is. “Heated Jacket Women Best Warm USB Rechargeable 2026” tells it you are gaming the field.

2. description

The description field is a one or two paragraph account of what the product is, who it is for, and what it is made of, matching the copy on the page. Shopify pulls it from your product description, which means thin descriptions produce thin schema. This is the field ChatGPT quotes when it recommends a product, and the same brief applies as when you write AI product descriptions that rank: facts first, then who it suits.

3. offers (price and priceCurrency)

The offers field holds the price and the currency, and those two values are the only required Offer properties in Google’s spec. The price has to be above zero and the currency has to be an ISO code like USD. Shopify handles both, including the strikethrough compare-at price during a sale.

Also, watch priceValidUntil if you run timed sales. Google reads it as the sale end date, so a stale value tells Google your sale ended last month while your page still shows the discount.

4. availability

The availability field says InStock, OutOfStock, PreOrder, or BackOrder, and it has to be right today, not last month. Google’s spec lists it as recommended, but a listing whose stock status disagrees with the page gets disapproved in Merchant Center, and ChatGPT’s feed spec makes it one of its nine required fields. Shopify updates it from inventory automatically, which is one reason to leave the theme’s block in charge.

5. brand

The brand field is your brand name as a Brand object, and it matters more for AI answers than for classic search. OpenAI’s product feed spec requires brand on every item and asks for it “as shown on the product page.” When a shopper asks ChatGPT for the best heated jacket for women, the brand field is how the answer ties your product page to the reviews and the press about it.

6. aggregateRating and review

The aggregateRating field carries the average star value and the count, and the review field carries the actual review text, and together they are the difference between a bare price in the results and a listing with stars. Google’s policy is that the ratings you mark up have to be the ratings a shopper sees on the page. Your review app either feeds these into the theme’s block or writes its own, and only the first option is safe.

Beyond the stars, the review text matters on its own, because models read it. If you want to know how much weight it carries, I break it down in do AI models read your reviews, and the answer is a reason to keep the review text in the block.

7. gtin or mpn

The gtin or mpn field is the product identifier, a barcode number or manufacturer part number, that lets Google match your listing to everything else it knows about that product. Google lists gtin, mpn, and sku as recommended. OpenAI marks gtin optional but takes 8, 12, 13, or 14 digit codes and keeps the leading zeros, which is a tell that it uses them for matching. Shopify has a barcode field on every variant, and most stores leave it empty.

So fill the barcode field on every variant, because it is the one schema fix that also fixes your Shopping feed. Google Merchant Center flags missing GTINs, and the same number closes both gaps at once.

Where Each Field Shows Up

Each of the seven fields shows up in a different place: Google’s rich results, Google Shopping, and ChatGPT’s product feed each read the same block for a different job. The table is the version I keep open during an audit.

FieldGoogle merchant listingChatGPT product feed
nameRequiredtitle, required
descriptionRecommendeddescription, required
offers (price, currency)Required, price above zeroprice with currency, required
availabilityRecommended, needed for Shoppingavailability, required
brandRecommendedbrand, required
aggregateRating, reviewRecommended, shows the starsNot in the feed, read from the page
gtin or mpnRecommendedgtin, optional but used for matching

Read the middle and right columns together, and six of the seven fields are required somewhere, while the seventh, ratings, is the one that decides the click. The feed and the schema pull from the same product data in Shopify, so a store that gets its product pages right has already done most of the work to get its products into ChatGPT Shopping, and the same block is what an AI agent reads in agentic commerce.

Why AI Surfaces Read Product Schema

AI surfaces read product schema because it is the one version of your page a model can trust without interpreting it, since the price, stock, and rating sit in labeled fields instead of buried in a paragraph. OpenAI built its shopping feed around the same fields.

Shopify says the same thing in its own guide to optimizing your store for AI, which tells merchants to add structured data and product attributes so AI systems can read the product page. A product page an AI can parse is a product page an AI can recommend.

Still, Google is careful about what schema is not. Its AI features documentation says you do not need new markup, AI text files, or any special schema.org structured data to appear in AI Overviews or AI Mode. The same page then tells you to make sure your structured data is valid and your Merchant Center feed is current. So schema is not the AI ranking lever, it is the confirmation layer, and machines reward pages they do not have to guess about.

Even so, the classic rich results payoff stands on its own. Google’s structured data case studies include Nestlé measuring 82% higher click-through on pages that show as rich results and Rotten Tomatoes seeing 25% more clicks on marked-up pages. Stars and a price in a search result is the closest thing to ad formatting an organic listing gets.

Then the newer surfaces read the same block. Google’s AI product carousels pull price and rating from the merchant listing pipeline, which is half the work of getting a product page into Google AI Overviews. ChatGPT found consensus between a #1 Google ranking, a Travel + Leisure pick, and a clean product page when it named the Fieldsheer Ventani Hybrid its top heated jacket for women. The schema did not win that on its own. It made sure nothing on the product page argued with the sources.

How to Check and Fix Your Product Schema

You check and fix your product schema in four steps that take about 20 minutes per quarter, and the three things that go wrong are missing fields, duplicate blocks, and values that drifted from the page, and each step catches one. A page with two broken blocks gets nothing from a new field, so keep the order.

  1. Run your top five product pages through Google’s Rich Results Test and note every error and warning under the Product result. Warnings are the recommended fields you skipped.
  2. Open the Merchant listings report in Search Console, which Google added in 2022 so stores could see product markup errors sitewide, and work the list from most pages affected down.
  3. View the page source and count the application/ld+json blocks that declare a Product. More than one means a theme and an app are both writing schema, so pick the theme’s block and turn the other off.
  4. Spot-check the values against the page. Price, stock, and rating should match what a shopper sees today. Google’s structured data policies treat markup that does not match the visible page as a violation.

After that, leave it alone until your products change. Schema is plumbing, and plumbing is done when it stops leaking. The ongoing work is keeping reviews flowing and inventory accurate, not adding more markup. If you would rather have the audit run for you, my AI search services start with exactly this pass on the top 20 products.

Product Schema Questions I Get Asked

Do you need a schema app on Shopify?

You do not need a schema app on most Shopify stores. Dawn and the themes built since 2021 already write a complete Product block, and a second app usually creates the duplicate problem it claims to fix. The exception is an older theme with no JSON-LD at all. Check the page source before you pay for one.

Does product schema help you rank?

Product schema does not help you rank in the classic sense, because Google has said structured data on its own is not a ranking factor, but it decides whether your listing shows stars and a price, and that decides the click. Ranking gets you on the page. The rich result gets you chosen once you are there. The same fields then feed the AI answers that increasingly happen before anyone sees a results page, which is why I treat schema as step one when a client wants ChatGPT to recommend their products.

Want Your Top 20 Products Checked?

Shopify product schema comes down to seven fields in one clean block per product page: name, description, price and currency, availability, brand, ratings and reviews, and a GTIN. Your theme almost certainly writes the block already, your apps are the most likely thing breaking it, and the same data earns rich results, feeds Google’s AI carousels, and matches the ChatGPT product feed. Complete it, deduplicate it, and check it quarterly.

If you would rather not own that checklist, I help Shopify brands stop renting customers from Meta and Google by building the search and AI visibility that keeps sending buyers after the ads turn off. Book a call and I will pull your top 20 product pages through the Rich Results Test and show you which of the seven fields are missing. If you want the fixes handled too, that is what my AI search services are for.

Sources:

SEO Consultant

Tyler Ziton is an SEO consultant and founder of Ziton Digital. He has driven $21M in annual organic revenue for 30+ brands including Park Seed, Flexfire, and Fieldsheer, and builds organic revenue systems for Shopify stores using SEO, AI search, and CRO.

Similar Posts